Automated Design and Discovery of Liquid Electrolytes for Lithium-Ion Batteries

The world requires an upgrade in battery performance before the ubiquitous electrification of transportation is feasible. But the novel materials that could unlock safer, more energy-dense batteries are difficult to discover. The battery material design process suffers from too much choice and frequent trade-offs, resulting in decades of research yielding only a handful of winning materials. 

New forms of laboratory automation hold promise for reducing the time and capital spent in research and development. A fully automated experiment can be coupled to a learning model to rapidly iterate on material designs without human involvement - “autonomous experimentation”. This thesis is the first attempt at automating the discovery of functional liquid electrolytes for batteries. Two instances of automated electrolyte experiments are presented - both featuring new hardware and software packaged into functioning test-stands. Three instances of automated optimization of liquid electrolytes for batteries are presented - one in the aqueous design space, and two in the non-aqueous space